Introduction

Chronic conditions constitute the bulk of the disease load, particularly in Western societies. The prevalence and consequences of such conditions are becoming better appreciated, although the intricacies of their pathogenesis and inter-relationships are largely not definitively determined. Major depressive disorder and osteoporosis are two pertinent examples of increasingly prevalent disorders that appear to be related. Several epidemiological studies have reported diminished bone mineral density among depressed men and women compared with controls (Reference Schweiger, Deuschle and Korner1–Reference Jacka, Pasco and Henry10), dysregulation of bone turnover (Reference Michelson, Stratakis and Hill3,Reference Yazici, Akinci, Sutcu and Ozcakar5,Reference Kahl, Greggersen and Rudolf8) and increased risk for fracture (Reference Whooley, Kip, Cauley, Ensrud, Nevitt and Browner11–Reference Mussolino13). Data from animal models show that stress-induced depression leads to altered rates of bone remodelling Reference Patterson-Buckendahl, Grindeland, Shakes, Morey-Holton and Cann(14) and loss of trabecular bone volume, trabecular number and connectivity Reference Yirmiya, Goshen and Bajayo(15). Although these studies are heterogeneous in design, they reveal a consistent association between depression and indices of osteoporosis. The nature of this association remains to be clarified, but a number of pathophysiological pathways are potential mediating mechanisms. Among these, oxidative stress stands out as a plausible candidate, given the fundamentality of oxidation physiology, and the newly emerging evidence for an intrinsic role of oxidative stress in both major depression and osteoporosis.

Oxidative stress

At higher concentrations, free radicals and non-radical reactive species are hazardous for living organisms and damage major cellular constituents. At lower concentrations, however, nitric oxide and reactive oxygen species (ROS, including superoxide, singlet oxygen, hydrogen peroxide and the highly reactive hydroxyl radical) play important roles as regulatory mediators in signalling processes Reference Droge(16,Reference Haddad17). Many ROS-mediated responses protect cells against exposure to oxidants and re-establish reduction/oxidation (redox) homeostasis. Excess ROS disrupt cellular redox status, leading to a net oxidative overbalance (‘oxidative stress’), which causes cellular dysfunction, and has been implicated in the pathogenesis of disease including cancer Reference Dreher and Junod(18), diabetes mellitus Reference Wolff(19), atherosclerosis Reference Alexander(20) and rheumatoid arthritis Reference Mapp, Grootveld and Blake(21). Progressive changes in free-radical-mediated regulatory processes may also play a role in the aetiology of both depression and osteoporosis and may, in part, mediate their association.

Oxidative stress and depression

Altered levels of antioxidative enzymes (Reference Bilici, Efe, Koroglu, Uydu, Bekaroglu and Deger22–Reference Herken, Gurel and Selek25) and the lipid peroxidation product malondialdehyde (Reference Bilici, Efe, Koroglu, Uydu, Bekaroglu and Deger22–Reference Sarandol, Sarandol, Eker, Erdinc, Vatansever and Kirli24) have been reported in major depression, especially in cases with melancholic features Reference Bilici, Efe, Koroglu, Uydu, Bekaroglu and Deger(22). Normalisation of oxidative stress markers with antidepressant treatment has been demonstrated (Reference Bilici, Efe, Koroglu, Uydu, Bekaroglu and Deger22,Reference Khanzode, Dakhale, Khanzode, Saoji and Palasodkar23,Reference Herken, Gurel and Selek25), and a positive correlation was reported between oxidative stress index values and severity of depression Reference Yanik, Erel and Katoi(26). These findings strengthen the hypothesis that oxidative stress is inherent in active depressive illness processes.

Further support for the relevance of oxidative stress behind depressive illness pathogenesis can be found in pharmacology studies. Established antidepressants have been shown to possess antioxidant properties in cellular and animal studies. For instance, the monoamine oxidase (MAO) inhibitor phenelzine has been observed using in vitrocell studies to modulate oxidative stress pathways via several mechanisms that included hydrogen peroxide scavenging and the prevention of MPP+-induced decrease in mitochondrial membrane potential, cytochrome crelease, formation of ROS and depletion of glutathione Reference Lee, Han and Lee(27). The reversible inhibitor of MAO-A, moclobemide, has been shown to significantly increase the survival of rat neuronal-astroglial cultures using in vitromodels of brain ischaemia, with decreased free radical formation through MAO-A inhibition being the hypothesised mechanism Reference Verleye, Steinschneider, Bernard and Gillardin(28). Venlafaxine reduced levels of oxidative stress in a rat model of depression (induced by chronic mild stress), as indexed by increased glutathione peroxidase activity and glutathione and vitamin C levels and by lower nitric oxide levels and lipid peroxidation in the brain Reference Eren, Naziroglu and Demirdas(29). Both venlafaxine and fluoxetine have been shown to modulate antioxidant proteins, among others, in a proteomic study Reference Khawaja, Xu, Liang and Barrett(30).

Oxidative stress and osteoporosis

Bone loss occurs as a result of imbalances in bone remodelling, whereby osteoclastic bone resorption exceeds osteoblastic bone formation Reference Riggs and Khosla(31). Adult ageing is accompanied by bone loss Reference Henry, Pasco, Nicholson, Seeman and Kotowicz(32), decreases in plasma antioxidant enzymes Reference Ceballos-Picot, Trivier, Nicole, Sinet and Thevenin(33) and reduced antioxidant intakes in the diet Reference Mertz(34). The rate of bone remodelling and negative bone balance increases after menopause in women, accelerating bone loss and trabecular and cortical thinning, and increasing bone porosity and fragility Reference Mosekilde(35).

There is compelling in vitroand in vivoevidence to suggest that oxidative stress is involved in bone cell biology. Antioxidant defences were found to fall in the bone marrow of female rats after ovariectomy, and the response was rapidly normalised by exogenous 17-β oestradiol and prevented by the antioxidants N-acetylcysteine (NAC) and ascorbate Reference Lean, Davies and Fuller(36). By contrast l-buthionine-(SR)-sulphoximine (BSO), a specific inhibitor of glutathione, the major intracellular antioxidant, caused substantial bone loss. Furthermore, depletion of antioxidants by BSO, like ovariectomy, caused bone loss through tumour necrosis factor-α (TNF-α) signalling, and in ovariectomised mice treated with soluble TNF-α receptors, antioxidant defences in bone remained low, despite inhibition of bone loss Reference Lean, Kirstein, Urry, Chambers and Fuller(37). The authors concluded that low levels of antioxidants in bone following ovariectomy were the cause of increased bone resorption.

The antioxidant vitamins are vitamins C and E and the β-carotene form of vitamin A. Guinea pig studies have shown that low vitamin C intake in growing animals is associated with high rates of bone turnover Reference Kipp, Grey, McElvain, Kimmel, Robinson and Lukert(38). Epidemiological studies have found a positive association between dietary vitamin C intake and bone mineral density in postmenopausal women (Reference Hall and Greendale39–Reference Wang, Luz Villa, Marcus and Kelsey41), and low dietary intakes of vitamins C and E have been linked to an increased risk of hip fracture in smokers Reference Melhus, Michaelsson, Holmberg, Wolk and Ljunghall(42). Diminished dietary and endogenous antioxidants have been reported among elderly osteoporotic women compared with age-matched controls Reference Maggio, Barabani and Pierandrei(43), with differences significant for dietary vitamins A, C and E, and for endogenous molecules with scavenging activity (specifically uric acid, superoxide dismutase and glutathione peroxidase). Antioxidant vitamin supplements have been associated with reduced rates of bone resorption in non-smoking postmenopausal women Reference Pasco, Henry, Wilkinson, Nicholson, Schneider and Kotowicz(44), and a recently reported randomised controlled trial showed progressively reduced bone resorption in early postmenopausal women treated with the potent orally administered antioxidant, NAC Reference Sanders, Kotowicz and Nicholson(45). Vitamin intake may, however, be associated with other risk factors.

It should be noted that excess dietary retinol intake has been identified as a risk factor for hip fracture and associated with accelerated bone loss Reference Melhus, Michaelsson and Kindmark(46,Reference Feskanich, Singh, Willett and Colditz47). It has been suggested that serum concentrations of vitamin A at both ends of the spectrum (high and low) increase the risk of hip fracture Reference Opotowsky and Bilezikian(48). The role of retinol, in addition to other forms of vitamin A (including β-carotene), remains unclear. However, taken in aggregate, the findings suggest a positive role for antioxidants in protecting against bone loss.

Oxidative stress connecting depression and osteoporosis

An initial hypothesis for the association between osteoporosis and depression related to hypercortisolaemia, which has been observed in major depression and is known to be detrimental to the skeleton Reference Seeman, Wahner, Offord, Kumar, Johnson and Riggs(49). However, this seems more suitable in explaining a unidirectional relationship and is unlikely to be the sole mechanism of the association between these two conditions, given the complexities of their aetiologies and pathogenic processes.

An advantage of examining pathogenesis from the oxidative stress perspective is the explanatory power afforded by its fundamental placement in the hierarchical aetiopathogenic framework, which extends from gross extrinsic influences to basic molecular biology. Hence, regardless of the initial insults responsible for the illness, oxidative stress may be a common trunk pathogenic pathway, which may become a self-perpetuating process that progressively induces disease states in susceptible individuals. Because of this basic role and its ubiquitous involvement in all bodily systems, oxidative stress stands out as a prime candidate mediator in the pathophysiology of comorbid diseases. As there is already evidence that establishes an active role for oxidative stress in the development of both major depression and osteoporosis, the extension of this role to a mediating one seems most plausible on both theoretical and empirical grounds.

It would of course be simplistic and unconvincing to assert that oxidative stress is the only mediating mechanism, when most diseases including major depression and osteoporosis involve multiple pathological paths. The capacity for the oxidative stress hypothesis to incorporate the action of other mechanisms through the confluence of oxidative pathways with other biopathways is one of its strengths. Immunological pathways provide a useful example for this discussion. For major depression and osteoporosis, each has individually accrued literature on immunological changes that seem intrinsically related to the disease process. In major depression, these include the acute-phase response, comprising raised haptoglobin, interleukin (IL)-6, complement factors and C-reactive protein (Reference Maes, Scharpe and Van Grootel50–Reference Hornig, Goodman, Kamoun and Amsterdam53), impaired lymphocyte proliferative response to mitogens and reduced lymphocyte phagocytosis Reference McAdams and Leonard(54). Elevated serum high-sensitivity C-reactive protein (hsCRP) has been reported in a large sample of women with osteopenia or osteoporosis and has been positively correlated with serum alkaline phosphatase levels, indicating states of accelerated bone turnover Reference Koh, Khang and Jung(55). Serum hsCRP has recently been identified as a prognostic marker of fracture (Reference Pasco, Kotowicz and Henry56–Reference Cauley, Danielson and Boudreau58). Various anti-inflammatory and proinflammatory cytokines including IL-1, IL-6, TNF-α and -β, transforming growth factor-α and -β, granulocyte-monocyte colony-stimulation factor, monocyte-macrophage-stimulating factor and prostaglandin E₂ are understood to be important in the normal bone remodelling process Reference Quinn and Gillespie(59,Reference Tanabe, Ito-Kato and Suzuki60). There is abundant evidence that cytokines and the inflammatory response are closely linked to oxidative stress states (Reference Lau, Shukitt-Hale and Joseph61–Reference Joseph, Shukitt-Hale, Casadesus and Fisher63), and this might provide a further nexus in the mediating role of oxidative stress for depression and osteoporosis.

Although oxidative stress is unlikely to be the sole mediator between major depression and osteoporosis, it offers an attractive hypothesis for further study for several reasons. Its omnipresent role in human physiology and its applicability to diverse, if not universal, pathological processes suggest that it may be comparatively dominant among different biological pathogenic mechanisms. Furthermore, it may have important practical implications in that antioxidants may have undiscovered preventive and therapeutic values. The oxidative stress hypothesis raises the possibility that sustaining antioxidant defences during ageing may be protective against these diseases. Antioxidant defences may be sustained through antioxidant intake, body habitus and lifestyle, and this field of research warrants further investigation. Nutritional epidemiology has already suggested that dietary antioxidants are crucially involved in the prevention of certain degenerative diseases, but at this time, the scientific data do not justify the use of antioxidants for risk reduction of depression or osteoporosis. We anticipate data for developing evidence-based recommendations for new methods of intervention using available tolerable, safe and affordable antioxidant approaches.

Conclusions

We have postulated and explored the concept that development of major depression and deterioration of the skeleton are both influenced by increased oxidative stress levels. Indeed, markers of oxidative stress within depression and osteoporosis may play a vital role in disease expression and suggest a mediating link between these disorders. In addition to pathological understanding, this hypothesis introduces opportunities to investigate novel treatment approaches, which may be of benefit to the outcomes of debilitating chronic diseases.